© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 37596
    Prev
    Next
    INDEX
    Explanations

    archive or archived content

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    adamkarvonen/qwen3-32b-saes/saes_Qwen_Qwen3-32B_batch_top_k/resid_post_layer_32
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    obb
    -0.10
    ened
    -0.10
    /or
    -0.10
     ÃŃch
    -0.10
    ality
    -0.09
    arker
    -0.09
    .appspot
    -0.08
    ities
    -0.08
    etur
    -0.08
    ly
    -0.08
    POSITIVE LOGITS
    äºĨä¸įå°ij
    0.11
    äºĨèĩªå·±çļĦ
    0.10
    edly
    0.10
    нем
    0.10
    SHOT
    0.10
    able
    0.09
    yard
    0.09
    лан
    0.09
    äºĨ大éĩıçļĦ
    0.09
    IRD
    0.09
    Activations Density 0.014%

    No Known Activations